• Journal of the Korean Statistical Society
• /
• 제23권1호
• /
• pp.167-178
• /
• 1994

To estimate coefficient matrix in autoregressive model, usually ordinary least squares estimator or unconditional maximum likelihood estimator is used. It is unknown that for univariate AR(p) model, unconditional maximum likelihood estimator gives better power property that ordinary least squares estimator in testing for unit root with mean estimated. When autoregressive model contains multiple unit roots and unconditional likelihood function is used to estimate coefficient matrix, the seperation of nonstationary part and stationary part of the eigen-values in the estimated coefficient matrix in the limit is developed. This asymptotic property may give an idea to test for multiple unit roots.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1998년도 제13차 학술회의논문집
• /
• pp.429-434
• /
• 1998

The previously developed control design methodology, EALQR(Eigenstructure Assignment/LQR), has better performance than that of conventional LQR or eigen-structure assignment. But it has a constraint for the weigting matrix in LQR, that is the weighting matrix could be indefinite for high-order systems. In this paper, the effects of the indefinite weighting matrix in EALQR on the Sequency domain properties are analyzed. The robustness criterion and quantitative frequency domain properties are also presented. Finally, the frequency do-main properties of EALQR has been analyzed by applying to a flight control system design example.

• 한국항해항만학회지
• /
• 제28권4호
• /
• pp.305-309
• /
• 2004

The problem for parameters estimation of the received signals impinging on array sensors has long been of great research Interest in a great variety of applications, such as radar, sonar, and land mobile communications systems. Conventional subspace-based algorithms, such as MUSIC and ESPRIT, require an extensive computation of inverse matrix and eigen-decomposition In this paper, we propose a new parameters estimation algorithm via nonlinear minimization, which is simplified computationally and estimates signal parameters simultaneously.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 추계학술대회 논문집 전력기술부문
• /
• pp.85-87
• /
• 2002

The paper describes the formulation of state matrix equations from the linearized multi-machine power system including network dynamics and the application of IEEE First Benchmark Model. The eigenvalues of IEEE First Benchmark Model are investigated not only by changing the compensation of series capacitance at no-load conditions, but also by varying the generator loading at fixed compensation of capacitance. In addition, the pure electrical self-excited mode is also examined by an eigen analysis and time domain simulation.

• Structural Engineering and Mechanics
• /
• 제24권6호
• /
• pp.683-698
• /
• 2006

The accurate dynamic analysis of concrete arch dams relies heavily on employing a three-dimensional semi-infinite fluid element. The usual method for calculating the impedance matrix of this fluid hyper-element is dependent on the solution of a complex eigen-value problem for each frequency. In the present study, an efficient procedure is proposed which simplifies this procedure amazingly, and results in great computational time saving. Moreover, the accuracy of this technique is examined thoroughly and it is concluded that efficient procedure is incredibly accurate under all practical conditions.

• 한국컴퓨터정보학회논문지
• /
• 제16권2호
• /
• pp.9-16
• /
• 2011

GPU는 저렴한 비용으로 쉽게 대규모 데이터 병렬성을 활용할 수 있는 장점을 갖고 있어 많은 고성능 컴퓨팅 응용 분야에서 사용되고 있는 추세다. 행렬의 고유벡터를 구하는 power method는 웹 페이지의 중요도를 계산하는 PageRank 알고리즘 등 여러 응용 분야에서 활용되고 있는 방법으로써, 본 연구에서는 power method를 GPU에서 병렬화하여 구현하였으며, 성능을 최적화하기 위한 개선 방법을 제시하였다. Power method는 행렬과 벡터의 곱셈 연산이 반복적으로 수행되며 GPU에서 쉽게 병렬화가 가능하다. 그러나, 고유벡터의 수렴 여부 판단을 위한 연산 등의 작업과 다음 곱셈을 위한 벡터 크기의 조정 등의 작업이 부가적으로 필요하며, 이러한 작업은 GPU 내의 커널 코드를 여러 차례 호출하고 불필요한 데이터 이동을 유발하는 문제점이 있다. 본 연구에서는 커널 호출 회수를 줄이고 스레드 배치를 최적함과 동시에 수렴 여부 판단을 위한 연산을 최적함으로써 power method의 성능을 향상시켰다.

• Earthquakes and Structures
• /
• 제6권6호
• /
• pp.647-664
• /
• 2014

A reduction of the response of irregular structures subjected to earthquake excitation by control devices equipped by suitable control algorithm is proposed in this paper. The control algorithm, which is used, is the pole placement one. A requirement of successful application of pole placement algorithm is a definition-selection of suitable poles (eigen-values) of controlled irregular structures. Based on these poles, the required action is calculated and applied to the irregular structure by means of control devices. The selection of poles of controlled irregular structure, is a critical issue for the success of the algorithm. The calculation of suitable poles of controlled irregular structure is proposed herein by the following procedure: a fictitious symmetrical structure is considered from the irregular structure, adding vertical elements, such as columns or shear walls, at any location where is necessary. Then, the eigen-values of symmetrical structure are calculated, and are forced to be the poles of irregular controlled structure. Based on these poles and additional damping, the new poles of the controlled irregular structure are calculated. By pole placement algorithm, the feedback matrix is obtained. Using this feedback matrix, control forces are calculated at any time during the earthquake, and are applied to the irregular structure by the control devices. This procedure results in making the controlled irregular structure to behave like a symmetrical one. This control strategy can be applied to one storey or to multi-storey irregular buildings. Furthermore, the numerical results were shown that with small amount of control force, a sufficient reduction of the response of irregular buildings is achieved.

• 한국음향학회지
• /
• 제38권2호
• /
• pp.177-183
• /
• 2019

본 논문에서는 수동 선배열 소나의 저주파 영역에서 수신된 표적 신호로부터 간섭신호를 분리해 내기 위해 독립성분분석 알고리즘을 적용하는 방안을 제안하고 기존 알고리즘들의 성능을 비교해 보았다. 저주파 대역 신호의 경우 비교적 넓은 방위로부터 수신되기 때문에 인접 빔 신호를 관측신호로 활용하여 독립성분분석을 수행할 수 있다. 신호분리에 사용한 독립성분분석 알고리즘은 FastICA(Fast Independent Component Analysis), NNMF (Non-negative Matrix Factorization), JADE (Joint Approximation Diagonalization of Eigen-matrices)이다. 실측 선배열 수동소나신호를 이용하여 독립성분분석을 수행한 결과 제안한 방법으로 간섭신호분리가 가능함을 확인하였으며, JADE 알고리즘의 신호 분리 성능이 가장 우수한 것으로 나타났다.

• 대한전기학회논문지:전력기술부문A
• /
• 제49권5호
• /
• pp.233-241
• /
• 2000

This paper presents a systematic design procedure of $H_{\infty}$ controller of TCSC for damping low frequency inter-area oscillations in large power systems. Sensitivities of the inter-area mode for changes in line susceptance are computed using the eigen-sensitivity theory of augmented system matrix and TCSC locations are selected using the line sensitivities. The reduced model required for designing a manageable-size $H_{\infty}$ controller is obtained using the reduced frequency domain system identification method and the various weighting functions are tuned systematically to provide a robust performance. The proposed $H_{\infty}$ controller proved to be very effective for damping the inter-area mode of the large KEPCO power system.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 하계종합학술대회 논문집
• /
• pp.687-690
• /
• 1999

Many high-resolution algorithms based on the eigen-decomposition analysis of observed covariance matrix, such as MVE, MUSIC, and EVM, have been proposed. However, the resolution of spectral estimates for these algorithms is severely degraded when Signal-to-Noise Ratio (SNR) is low and arrival angles of signal are close to each other. And EVM and MUSIC is powerful for the characteristic of SNR. But have the limitation that the number of signals presented is known. While MVE is bad the characteristic of SNR. In this study, we propose a modified MVE to enhance the resolution for Direction-Of-Arrival (DOA) estimation of underwater acoustic signal. This is to remove the limitation that existing algorithms should know the information for the number of signals. Because the algorithms founded on the eigen value estimate DOA with only the noise subspace, they have the high-resolution characteristic. And then, with the method reducing the effect of the signal subspace, we are to reduce the degradation because of complementary relationship between the signal subspace and the noise subspace. This paper, with using the simulation data, we have estimated the proposed algorithms, compared it with other high-resolution algorithms. The simulation results show that the modified MVE proposed is accurate and has a better resolution even though SNR is low, under the same condition.